Author Topic: How to read the fuse charts (I^2t)  (Read 8840 times)

0 Members and 1 Guest are viewing this topic.

Offline T_guttataTopic starter

  • Regular Contributor
  • *
  • Posts: 134
  • Country: 00
How to read the fuse charts (I^2t)
« on: August 03, 2021, 06:05:17 pm »
Hi

Could anyone help me out on how to read the data sheet of fuses:

https://www.bourns.com/docs/product-datasheets/sf-2410hi-t.pdf?sfvrsn=4d8a79f6_6

A fuse with a rated current of 0.5A has an I^2t of 0.495. How can I determine the maximum peak currents (and time) the fuse will alow without triggering? What will the max current be for a pulse width of 10ms? 50ms?

Thanks for your help.
 

Offline cortex_m0

  • Regular Contributor
  • *
  • Posts: 114
  • Country: us
Re: How to read the fuse charts (I^2t)
« Reply #1 on: August 04, 2021, 02:59:03 am »
The i2t rating is essentially defining a curve between current and time to blow.

i2t = 0.495

So you can pick i, or t, and solve. You proposed t = 50ms in your question.
i2(.05 s) = 0.495
i = 3.14A

If you want to validate that, go to the charts on page 2. Find the "Average Pre-Arcing Time vs Current" chart. Follow the curve for the 500mA part to its crossing of 3.14A. You'll find that the "pre-arcing time" is approx 50ms.

Note that fuses are far from ideal components. Dave did a video recently on what happens if you run a fuse slightly higher than its rated current continuously.
 

Offline T_guttataTopic starter

  • Regular Contributor
  • *
  • Posts: 134
  • Country: 00
Re: How to read the fuse charts (I^2t)
« Reply #2 on: August 04, 2021, 04:47:23 am »
But if I solve for the rated current of 0.5A, I get only t~2s?
 

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 7858
  • Country: us
Re: How to read the fuse charts (I^2t)
« Reply #3 on: August 04, 2021, 05:17:06 am »
But if I solve for the rated current of 0.5A, I get only t~2s?

You'll see that the **** note tells you that the I2T number is at 10X rated current, so 5 amps.  The I2T equation will roughly hold at currents higher than this, but since the fuse dissipates heat it obviously can't work for low currents.  If you used your method and solved for 50mA, you'd get t~200 seconds, and clearly that isn't right...
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Offline golden_labels

  • Super Contributor
  • ***
  • Posts: 1209
  • Country: pl
Re: How to read the fuse charts (I^2t)
« Reply #4 on: August 04, 2021, 09:26:54 am »
10ms is outside of the specs for this fuse. Of course there may be some current which will cause it to blow in 10ms, but it’s not guaranteed by the datasheet. So I will skip that.

Fuses are not precision devices. Dave’s video shows that experimentally, but you can derive that yourself from the clearing time table on page 1. See the attachment for how you should visualize the table. Note that the time (vertical) axis is logarithmic. The green indicators are values from the table. The dark reddish nebulous blob is a ballpark estimate of where the fuse blows according to the datasheet. A bit over 100% the rated current is between a minute and a week, I would say. A bit over ampere for a 500mA fuse and you can’t tell if it blows in half a second or a minute.

50ms, you ask? Find the 50ms mark on the vertical axis, look to the right. My eye tells me it’s somewhere between 3.5× to 8× rated current (1.75–4A). Though above 8× there be dragons and those dragons seem to also sit comfortably around the 50ms line.

Now, you asked about I²t. Fuses are not reacting to current. Fuses are reacting to heat. The heat comes from energy dumped into them. As it happens, the power lost in a fuse is more or less proportional to the current it passes and the total energy dumped into it is proportional to the integral of power over time. The energy will be the same no matter if it’s low power over long time or much power over short time. And, since power is proportional to current through the fuse, instead of using joules, A²s are used for convenience, but expressing the very same concept.

At this point you may wish to calculate I²t for e.g. 50ms overcurrent of 2A. You get (2A)²·0.05s = 0.2A²s, which is within the allowed range. But then also (0.49A)²·1000s gives 240A²s, which is clearly above the rated I²t and that is a contradiction, as it’s 490mA — below fuse’s rated current. Here comes another factor. The fuse can cool down. If it cools down fast enough, it will not blow. Below the rated 500mA it simply dumps heat at a rate high enough that it can remove heat produced by the current, leading to inifnitely long operation. The precise physical models behind that are Complicated™. But fortunately you don’t need to calculate it. On page 2 you have a I2t vs time chart. Calculate I²t of the circuit, find it in the chart, follow the vertical line. You are allowed to dump the amount of energy proportional to that I²t as long as you do that over time expressed on the horizontal axis, if your vertical line is below the curve for a fuse. Our 240A²s for 490mA over 1000s is just a bit under the cuve for the 500mA model. Hence the fuse can handle that. It may feel weird to first multiply by time and then effectively divide by it, instead of simply having a chart for current, but remember that what I²t expresses is total energy dumped, not simply constant current over time. The shape of the current surge may vary and I²t must be calculated for that particular situation.

Then there come three more issues. First: the chart indicates the mean value. Fuses are not precise devices. Being just under the curve on that chart is not safe. Have some safety margin. Second: that chart tells you in which conditions the fuse is expected to survive surges. Not when it is expected to blow. It may as well survive way more than indicated in the chart. Third: the chart is for the time it takes for the fuse to start melting. Unlike the table on page 1, which gives total time the fuse need to completely break the circuit, this chart does not take into account potential arcing.

In othe words: the I²t charts and values indicate the safe operating area for the fuse in temporary, expected overcurrent event. They do not provide information on how fast fuse blows. Which is why you can’t derive answer for your question from them.
« Last Edit: August 04, 2021, 09:53:40 am by golden_labels »
People imagine AI as T1000. What we got so far is glorified T9.
 
The following users thanked this post: tinfever

Offline Nusa

  • Super Contributor
  • ***
  • Posts: 2416
  • Country: us
Re: How to read the fuse charts (I^2t)
« Reply #5 on: August 04, 2021, 11:55:27 am »
Don't forget that all those numbers are usually at 25C. Temperature changes the equation. The temperature inside the box the fuse is located is likely above ambient! And ambient isn't a fixed value either.

I've actually run into cases where the fuse was specced too close to the line, but it worked most of the time. But once a really hot day happened, it would blow after a couple hours.
 
The following users thanked this post: SeanB

Offline T_guttataTopic starter

  • Regular Contributor
  • *
  • Posts: 134
  • Country: 00
Re: How to read the fuse charts (I^2t)
« Reply #6 on: August 04, 2021, 06:45:30 pm »
Thanks for your help. Well, in my case I need a fuse, which will allow high current pulses in the range of ms (5-10A) but will have the lowest continuous rating possible. Thats why I thought I need a slow fuse. Is it the correct fuse type, or are there alternatives?
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16284
  • Country: za
Re: How to read the fuse charts (I^2t)
« Reply #7 on: August 04, 2021, 08:12:39 pm »
Don't forget that all those numbers are usually at 25C. Temperature changes the equation. The temperature inside the box the fuse is located is likely above ambient! And ambient isn't a fixed value either.

I've actually run into cases where the fuse was specced too close to the line, but it worked most of the time. But once a really hot day happened, it would blow after a couple hours.

Had the same with a breaker, it was running at slightly above current rating, but if the load was applied for less than a certain time, it would not trip. Exceed this time by 5 seconds and it would reliably trip. I now have, forever as part of a installation book, a laminated piece of cardboard, that I grabbed from one of the boxes, that proposed a modification. Signed off by a PE at their local office as an approved modification to the site itself.  This is for a lift (elevator for those in the USA), where it would trip on overcurrent if you went from ground floor to the top, but not if you did a single stop anywhere short of that, just that 5 seconds of extra running would trip the breaker, with the car 1m short of the landing. Modification was to move 1 wire to another terminal on the 3 phase breaker, and add a single wire link, as they were using a 3 phase approved breaker with the one pole as part of the shaft safety circuit, instead of a dedicated and expensive overcurrent trip, that is notoriously more prone to failure, and is also not going to work without running the current through all 3 legs, to defeat the phase loss detection.
 

Offline golden_labels

  • Super Contributor
  • ***
  • Posts: 1209
  • Country: pl
Re: How to read the fuse charts (I^2t)
« Reply #8 on: August 04, 2021, 08:20:13 pm »
T_guttata: what’s the nature of those pulses? Is that a single event or are they repetitive?

A slow fuse is the right choice. But if it’s for example inrush current surge, you may also think about limiting the current in the first place by the means of putting an NTC in series with it.

People imagine AI as T1000. What we got so far is glorified T9.
 

Offline T_guttataTopic starter

  • Regular Contributor
  • *
  • Posts: 134
  • Country: 00
Re: How to read the fuse charts (I^2t)
« Reply #9 on: August 04, 2021, 09:06:45 pm »
It's quickly charging a cap. It will be repetitive, but with a very low duty cycle (<1%) I don't want to limit current too much, because I want to charge the cap fast. Currently it's not current controlled, therefore there will be a high peak, but it decreases rapidly.
 

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 7858
  • Country: us
Re: How to read the fuse charts (I^2t)
« Reply #10 on: August 04, 2021, 09:26:46 pm »
It's quickly charging a cap. It will be repetitive, but with a very low duty cycle (<1%) I don't want to limit current too much, because I want to charge the cap fast. Currently it's not current controlled, therefore there will be a high peak, but it decreases rapidly.

What you propose would be outside the data on the datasheet.  You can try extrapolating at your own risk.  Just do the math and realize that if you have a current of 20X the rating, you are generating 400X the heat.  So at a 1% duty cycle, you are generating heat equivalent to 4X the rating, or about the same as a continuous 1 amp would generate. 
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Offline T_guttataTopic starter

  • Regular Contributor
  • *
  • Posts: 134
  • Country: 00
Re: How to read the fuse charts (I^2t)
« Reply #11 on: August 05, 2021, 04:59:20 pm »
It's not that simple, because of the shape of the charge curve of a capacitor.

According to http://www.olino.org/blog/us/articles/2006/11/22/charge-efficiency-capacitor, charging a capacitor with a constant voltage source has an efficiency of 50% and is independant on the ESR.

If I take a 1000 uF cap and charge it to 30V, 1/2CV^2 = 0.45J is dumped into the series resistor.

According to the datasheet of the fuse, 0.375A is the rated current. With the fuse resistance of 0.6208 Ohm, heat dissipation inside the fuse will be 0.375*0.375*0.6208=0.087 W.

One charging cycle of the cap every 5.6s (0.18Hz) will generate the same amount of heat inside the fuse as the rated current. Is this correct?
 

Offline Alti

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: 00
Re: How to read the fuse charts (I^2t)
« Reply #12 on: August 05, 2021, 07:36:21 pm »
Quote
Is this correct?
Not really because the resistance of the link varies significantly, especially when it is getting red hot.
 

Offline T_guttataTopic starter

  • Regular Contributor
  • *
  • Posts: 134
  • Country: 00
Re: How to read the fuse charts (I^2t)
« Reply #13 on: August 05, 2021, 07:47:16 pm »
If the fuse is used at rated current, it will certainly not get red hot?
 

Offline Alti

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: 00
Re: How to read the fuse charts (I^2t)
« Reply #14 on: August 05, 2021, 08:48:33 pm »
If the fuse is used at rated current, it will certainly not get red hot?

Quote from: Clearing Time Characteristics for Series
100 % of Current Rating -> 4 hours, min
If it melts at rated current then what temperature can that be at?
 

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 7858
  • Country: us
Re: How to read the fuse charts (I^2t)
« Reply #15 on: August 05, 2021, 11:02:18 pm »
One charging cycle of the cap every 5.6s (0.18Hz) will generate the same amount of heat inside the fuse as the rated current. Is this correct?

Your idea to do an more detailed modeling is the right way to go, but you don't have enough information.  If the resistance of the fuse is constant and is what you say it is, and the capacitor has negligible ESR and the voltage source has a very low impedance, then this is correct.  However, the peak current will be over 50 amps and I would guess that the fuse would blow on the first pulse.  The rule that charging a capacitor dissipates half the energy in series resistance is true, but it includes ALL of the series resistance.  If you really need to use such a small fuse, you should use a larger series resistance, whatever your intended use can tolerate.  10 to 100 ohms maybe?
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Offline T_guttataTopic starter

  • Regular Contributor
  • *
  • Posts: 134
  • Country: 00
Re: How to read the fuse charts (I^2t)
« Reply #16 on: August 06, 2021, 07:05:58 am »
Well, @ 10 Ohm charging of the cap takes a bit too long. But I could restrict the initial current to 10A. Actually a constant current source would be better for charging a cap anyway. Then you don't have huge peaks, and efficiency can bi >50%. But I could not find a boost converter yet which is current controlled. Designing my own is not simple as I'm not an electrical engineer.

I will get some fuses see what happens :-)
 

Offline lovatt

  • Newbie
  • Posts: 1
  • Country: au
Re: How to read the fuse charts (I^2t)
« Reply #17 on: March 01, 2023, 04:52:46 am »
The full formula for the time to fuse is:

(ki2t / ki2) ln(i2 / ((i - ki)(i + ki)))

Where ki2t is the i2t rating and ki current rating of the fuse.

The formula is derived for a constant current, if the fuse time of a repetitive waveform is required then using the RMS will be approximately correct provided that the period is a lot less, say factor of 4 less, than the time constant, ki2t / ki2.

Note
  • The graphs given in the fuse data are often more accurate than the above formula.
  • The above formula tends to ki2t / ki2 as i tends to infinity.
.
 

Offline Vovk_Z

  • Super Contributor
  • ***
  • Posts: 1417
  • Country: ua
Re: How to read the fuse charts (I^2t)
« Reply #18 on: March 01, 2023, 12:55:25 pm »
But if I solve for the rated current of 0.5A, I get only t~2s?
I^2 rating is a rating to blow, so it is not about rated current but about higher-than-rated current.
 

Offline Vovk_Z

  • Super Contributor
  • ***
  • Posts: 1417
  • Country: ua
Re: How to read the fuse charts (I^2t)
« Reply #19 on: March 01, 2023, 12:59:24 pm »
If the fuse is used at rated current, it will certainly not get red hot?
No, rated current is the current at which a protective device (all of them typically) has to conduct current and doesn't blow so it mustn't be red hot.
 

Offline Vovk_Z

  • Super Contributor
  • ***
  • Posts: 1417
  • Country: ua
Re: How to read the fuse charts (I^2t)
« Reply #20 on: March 01, 2023, 01:01:56 pm »
If it melts at rated current then what temperature can that be at?
It mustn't melt at rated current until ambient temperature is 20-25 C.
 

Online Siwastaja

  • Super Contributor
  • ***
  • Posts: 8172
  • Country: fi
Re: How to read the fuse charts (I^2t)
« Reply #21 on: March 01, 2023, 03:38:10 pm »
If fast capacitor charge is a priority, then you just go for a switch mode constant current circuit. At the same time, you'll get the efficiency up from 50%.

Although, if the cap is rather small (such as 30V, 1000uF), then it's probably not worth the trouble. One middle ground solution is to provide two switched charge paths instead of one, with one having more series resistance than the other. Switch on the more resistive path first, wait until voltage is at maybe 50%, then switch on the second path. Efficiency is still just 50%, but now you have some control over the current without making the last drops take forever.
 

Offline golden_labels

  • Super Contributor
  • ***
  • Posts: 1209
  • Country: pl
Re: How to read the fuse charts (I^2t)
« Reply #22 on: March 01, 2023, 11:34:36 pm »
You folks do realize, you are answering an already answered 2021 thread? :)
People imagine AI as T1000. What we got so far is glorified T9.
 

Online Siwastaja

  • Super Contributor
  • ***
  • Posts: 8172
  • Country: fi
Re: How to read the fuse charts (I^2t)
« Reply #23 on: March 02, 2023, 06:58:18 am »
You folks do realize, you are answering an already answered 2021 thread? :)

No we don't. This happens all the time:
* Spambot bypasses poor registration security
* Spambot picks random old thread and posts something
* Only the spambot sees the "warning: you are replying to an old thread"
* Others miss this fact since the warning is only shown once

The root fix would be to increase security of forum registration (better captchas, etc.) Another improvement would be to make that "warning: you are replying to an old thread" message appear for 3-4, maybe 5 consecutive messages, after which the thread would be considered active again.
 
The following users thanked this post: Per Hansson

Offline Alti

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: 00
Re: How to read the fuse charts (I^2t)
« Reply #24 on: March 05, 2023, 12:18:15 am »
If the fuse is used at rated current, it will certainly not get red hot?

Quote from: Clearing Time Characteristics for Series
100 % of Current Rating -> 4 hours, min
If it melts at rated current then what temperature can that be at?

If it melts at rated current then what temperature can that be at?
It mustn't melt at rated current until ambient temperature is 20-25 C.
Hi. Did you bother opening the pdf from opening post?

First row in in "Clearing Time Characteristics" at ambient temperature and rated current, shows minimum clearing time as 4 hours, while maximum clearing time is not specified (it is not being tested).

Your response: this fuse must not melt at rated current.

Then contact Bourns and tell them they make lousy fuses as some burn in fifth hour at rated current.

I am curious what would have been their answer.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf